Building and Environment 96 (2016) 170e177

Contents lists available at ScienceDirect

Building and Environment

journal homepage: www.elsevier.com/locate/buildenv

A review on the properties of cellulose fibre insulation

Pablo Lopez Hurtado a, b, d, *, Antoine Rouilly a, b, Virginie Vandenbossche a, b,
Christine Raynaud a, c
a
University of Toulouse INP-ENSIACET, LCA (Laboratoire de Chimie Agro-industrielle), 31030, Toulouse, France
b
INRA, UMR 1010 CAI, 31030, Toulouse, France
c
CATAR-CRITT Agroressources, 31077, Toulouse, France
d
Greenbuilding, 75001, Paris, France

a r t i c l e i n f o a b s t r a c t

Article history: The building sector is constantly innovating in its use of materials with regards to sustainability. There is
Received 7 July 2015 a need to use cost effective, environmentally friendly materials and technologies which lessen the impact
Received in revised form of a construction in terms of its use of non-renewable resources and energy consumption. Cellulose fibre
28 August 2015
insulation is an eco-friendly thermal insulation material made from recycled paper fibres. It offers good
Accepted 21 September 2015
thermal properties and has a low embodied energy. However due to lack of expertise in its application
Available online 17 November 2015
and properties, cellulose insulation is not widely used in comparison to more traditional insulation
materials. The present paper reviews the available research on cellulose fibre insulation, its manufacture,
Keywords:
Cellulose insulation
installation, and performance. The paper focuses the physical properties of cellulose insulation, the
Recycled paper environmental factors that affect these properties, and possible means of future innovation.
Thermal conductivity © 2015 Elsevier Ltd. All rights reserved.
Settling
Moisture uptake
Borate
Fire resistance

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
2. A background on CFI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
2.1. Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
2.2. Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
2.3. Installation of CFI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3. Properties of CFI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.1. Density and settling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.2. Thermal properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
3.3. Moisture properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
3.4. Air infiltration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
3.5. Fire properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
3.6. Fungal development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
3.7. Life cycle analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

* Corresponding author. Present adress: Laboratoire de chimie agroindustrielle,

e Emile Monso, 31030, Toulouse, France.
ENSIACET, 4 Alle
E-mail address: pablo.lopezhurtado@ensiacet.fr (P. Lopez Hurtado).

http://dx.doi.org/10.1016/j.buildenv.2015.09.031
0360-1323/© 2015 Elsevier Ltd. All rights reserved.
 P. Lopez Hurtado et al. / Building and Environment 96 (2016) 170e177
Table 1
Embodied energy per kg of different insulation materials, data from Ref. [18].
Insulation material Embodied energy (MJ/kg)
General Insulation 45
Cellular Glass 27
Cellulose 0.94e3.3
Cork 4 fabrication and installation are presented. The available research on
Fibreglass (Glasswool) 28
Flax (Insulation) 39.5
Mineral wool 16.6
Paper wool 20.17
Rockwool 16.8
Woodwool (loose) 10.8
Woodwool (Board) 20
Wool (Recycled) 20.9
1. Introduction
Energy efficiency in buildings is an important factor in
contributing to the reduction of greenhouse gas emissions The
building and construction sector accounts for 30%e40% of world-
wide energy consumption [47], with a large part belonging to the
need to heat and cool buildings. It is with that in mind that many
countries are looking to improve the energy efficiency of buildings
with better insulation materials and technologies applied to the
building envelope, with directives such as the European directive
2010/31/EU which states that new constructions in 2020 will have
to consume 'nearly zero-energy' [14].
The main role of thermal insulation materials in a building en-
velope are to prevent heat loss and provide thermal comfort for a
building's interior. The factor that characterizes an insulation ma-
terial's effectiveness is its thermal conductivity l (measured in W/
mK). The lower a material's thermal conductivity, the more effec-
tive it is as an insulator, thus requiring a thinner layer to provide the
same interior temperature. Traditional insulation materials include
glass fibre, stone wool, expanded polystyrene, and polyurethane
foam. While these materials are efficient in maintaining thermal
comfort to a buildings interior, they are made with non-renewable
resources and have a high embodied energy. Consequently, there is
an increasing interest for alternative insulating materials that come
from renewable or recycled fibres. Natural fibres such as jute, flax
and hemp have shown to be suitable alternatives to mineral insu-
lation and are the subject of numerous research projects [23].
One such material is cellulose fibre insulation (CFI). Comprising
mostly of recycled paper fibres, cellulose is increasing in popularity
due to its eco-friendly nature and favourable thermal and acoustic
properties. Even amongst other insulation materials CFI presents
some of the lowest embodied energy per kg of material, as is shown
in Table 1 [18]. Despite growing interest, Cellulose and other natural
insulation materials still only represent a low percentage of total
European market share [29]. This is partly due to the fact that
cellulosic fibres, while having favourable thermal properties, still
have some disadvantages compared to traditional fibres. Some
factors such as its high hygroscopisity, potential for combustibility
and for fungal growth can limit CFI from having a more widespread
usage in construction and renovation projects. Proper knowledge of
these limits, their causes and their effect on the properties of the
insulation material are necessary to ensure that sustainable
Table 2
Average component proportions of newsprint and office paper, [50].
Cellulose % Hemicellulose %
Office paper 67.4 13
Newsprint 48.3 18.1
171
materials such as CFI become more common in the building sector
and thus help contribute to the reduction of the environmental
impact of construction and renovation projects.
The aim of this paper is to review the available information
regarding cellulose fibre insulation (CFI). First the general context
on CFI is given, including its background and main methods of
the properties of CFI is exposed, as well the different conditions that
affect these properties. Finally the paper comments on possibilities
for future investigation on the material and improvements of its
properties.
2. A background on CFI
2.1. Composition
Cellulose fibre insulation is mainly composed of ground paper
fibres treated with inorganic additives that act as fire retardants
and mould growth inhibitors. Its consistency is similar to that of
cotton wool. The source material for the cellulose fibres are usually
recycled newspaper, coming from either unsold or recovered pa-
pers. Newsprint is generally manufactured by mechanical pulp.
Recycled newsprint or chemical pulp could also be incorporated. As
with most lignocellulosic fibres, newsprint is comprised of a mix of
cellulose hemicelluloses and lignin. Unlike chemical pulping, me-
chanical pulping results in little removal of lignin content. Mineral
and organic additives, such as kaolins, china clay or cationic starch
are also incorporated into the paper pulp in order to improve such
properties as paper opacity, moisture retention, and strength. The
inks typically used in the paper are produced from inorganic car-
bons, with the chromatic inks coming from organic pigments. The
average proportions of the main components in newsprint and
office paper (chemical pulp) are presented in Table 2 [50].
2.2. Production
As a final product, cellulose insulation can come in two forms: as
a prefabricated panel, in which the cellulose fibres are moulded
with polyester or a similar binder or, more commonly, the loose
fibres are sold in bulk form to be manually applied on attics, ceil-
ings, or walls. The first use of cellulose fibre as an insulation ma-
terial can be traced back to 1919 in Canada [40], but it was until the
1950s that commercial cellulose insulation products became
commercially available in the US, where it was mostly used for attic
retrofitting. CFI surged in popularity in the US in the 1970s due to
an increased interest in energy performance following the Amer-
ican oil embargo of 1973.
In a typical production process of CFI (see Fig. 1), newspaper
arrives in bulk to the manufacturer and is then sorted to remove
any foreign objects. Items such as clips and plastics are removed,
but also low quality or humid paper is also sorted. The newsprint
passes through a feeding conveyor (1) then is torn to smaller pieces
that are between 2 and 4 cm in diameter in a shredder (2). The
fibres then pass through a cyclone separator (3) in order to remove
any remaining staples or other metallic elements. The fines from
the shredded paper are blown through a filtering unit (4). The
material then goes through a fiberizer (5) which uses high
Lignin % Extractives % Proteins % Ash %
0.93 0.7 0.31 11.6
22.1 1.6 0.44 2
172 P. Lopez Hurtado et al. / Building and Environment 96 (2016) 170e177
Fig. 1. Example of a typical manufacturing process of CFI, adapted with permission
from Ref. [24].
Fig. 2. Microscopic scan of cellulose fibres, 10X magnification.
pressured air to reduce the paper into low-density cotton-like
flakes, as shown in Fig. 2. It is in this stage that the powdered ad-
ditives are dispersed and mixed with the fibres. The additives used
are typically a mix of borax and boric acid, with a dose of around
15%e20% of the mass of cellulose fibres. A second cyclone separator
(6) then removes the fines created from the fiberizer. In the final
stage of the process, the fibres are filled in bags and then me-
chanically compacted (7) into 3 times its normal density (around
130 kg/m3), in order to reduce transport costs. The bags are then
weighed (8) and bundled into pallets and transported to supplier or
directly to construction sites.
2.3. Installation of CFI
There exist two main methods of applying CFI. Depending on
the desired properties, CFI can either be installed via the “loose fill”
or the “wet spray” method.
In the loose fill method the cellulose fibres are installed with
specific pneumatic blowing equipment. The compacted cellu-
lose is fed to the blower which separates the fibres which then
pass through the blowing system. The CFI is then delivered via
air pressure into closed wall/roof cavities or attics through a
hose. When cellulose is installed as “dense pack”, sheets of
netting are put in place over wall cavities. The cellulose is then
blown into the cavities between studs at a higher density than
loose fill, with the netting supporting the fibres. One of the
disadvantages of this method is that settling of the material
may occur over time, which decreases the insulation's effec-
tiveness, forming voids that cause thermal bridging in a
building envelope [7].
The wet spray technique is mostly used in open-wall wood
cavities separated by studs. It uses the same blowing equipment as
with loose-fill CFI, but a separate pump is used to spray water
simultaneously as the material is being blown with the cellulose in
order to improve the adherence of the fibres. After projection the
excess material is removed via a motorized wall scrubber and the
excess moist material is reintegrated in the blower. The water/CFI
mass ratio used in this process is typically around 40%e60%. Ad-
hesives, either mixed with the water or dispersed within the fibres
could also be used [8]. The main disadvantage of this method is that
drying times may vary, depending on the thickness and ambient
conditions of installation [37]. A variant of the wet spray method is
known as “stabilized” cellulose where a smaller dosage of water
(less than 20% in mass) is used to prevent dust and settling in
horizontal applications.
3. Properties of CFI
3.1. Density and settling
When dealing with loose fibres as an insulating material, it is
important to distinguish between the “blown” density and the
“design” density of the fibres. The blown density is the declared
density after installation in vertical or horizontal applications, and
the design density (which takes settling into account) is deter-
mined via impact testing and/or cyclic humidity testing. Impact
testing consists of subjecting the loose cellulose samples to a series
of vibrations. In cyclic humidity tests, the samples are subjected to
periodic variations of relative humidity [1]. One of the first studies
regarding the settling of CFI was done by [2]. Their study found an
average blown density of 34.8 kg/m3 for horizontal applications.
The average loss in thickness from settling was 21.5% wherein 10.5%
was from drop impact tests and 11% was from cyclic humidity
testing. The design density can be then calculated using by multi-
plying a factor which takes into account both types of settling:
Dd ¼ ð100=ð100  SÞÞDi (1)
Where Dd is the design density, Di is the installed density, and S is
the sum of both the settling from drop impact tests and cyclic
humidity testing.
The previous values give a design density factor of 1.27Di, thus
an average design density of 44.4 kg/m3 for horizontal applications.
It was also found that the dosage of fire additives increases density
linearly, although the type of additive or mix thereof has little in-
fluence on final density. A survey of 38 houses in six Canadian cities
[52] found the actual settling density, a year after installation, to be
averaged to 11.1%, with a range of 8.3%. The study suggests that the
blown density measured in laboratory be first multiplied by a factor
of 1.074 to account for differences between lab and building site
measurements, and then calculated with Eq. (1) using an average
settling of 11.1%.
For horizontal applications the compressibility of loose-fill CFI
can make its density vary widely. One early study by , [3] shows
installed density varying between 50 and 90 kg/m3. It was rec-
ommended to increase density by 10% after filling the wall cavity
in order to prevent settling, with a minimum density of 57 kg/m3.
A series of works by Rasmussen [31e34] have produced an
approach which allows to analytically determine the optimal
installed density of loose fill CFI that prevents settling in wood
frame walls. The method takes into account the dynamic me-
chanical behaviour of a typical insulated wall cavity that is sub-
jected to a cyclical variation in humidity in order to determine the
density required for the fibres to lose volume. The volume stable
 P. Lopez Hurtado et al. / Building and Environment 96 (2016) 170e177 173

was not significant (1%e3% increase). The study was done on facade
elements with 285 mm loose-fill CFI equipped with heat flow
meters and moisture measuring dowels. Heat flow measurements
were made according to the DS 418 standard. Sandberg [38]
developed three approaches to determine thermal conductivity as
a function of water absorption using moisture content profiles of
cellulose insulation. Measurements were made on 164 mm thick
loose fill CFI samples on 600 mm  600 mm frames, following the
ISO 8301 and ISO DIS10-051 standards. Computer simulations used
the following relation with regards to the thermal conductivity of
cellulose:

l ¼ 0:037 þ 0:00020 wðW=mKÞ: (2)

Where w is the mass of water per unit volume of cellulose kg/m3.

Fig. 3. Calculated minimal density for settling prevention of loose fill CFI in a wall The calculated results were in agreement with sample
under static humidity conditions (50% and 80% RH), as a function installed thickness, measurements.
top left shows the dimensions of the wall cavity [33]. Talukdar et al. [44] determined a polynomial function to
describe the relation between moisture and thermal conductivity
by curve fitting values measured by a heat flow meter apparatus
density of CFI was determined through the study of the creep, according to ASTM standard C518 on cellulose at different relative
coefficient of friction, and horizontal stress ratio testing of loose humidity conditions. Measurement temperatures were at 10  C and
fibres. As an example, the minimum density to prevent settling 350C, with an average temperature of 22.5  C
with CFI a 2.4 m tall, 0.1 m thick and 1 m wide gypsum wall at  
25  C and 50% RH was found to be 48 kg/m3. This value increases leff ¼ a þ b4 þ c41:5 þ d expð  4Þ (3)
linearly with wall thickness and relative humidity (Fig. 3). Dy-
namic conditions were also tested, where humidity varied from
where a ¼ 0.092482655, b ¼ 0.15480621, c ¼ 0.066517733 and
50% to 80%. In this case a 2.3 m high, 0.198 m thick and 0.495 m
d ¼ 0.1296168.
wide gypsum board cavity was calculated to require a density of
The only research that studied changes in thermal conductivity
62.3 kg/m3 to prevent settling. This was later confirmed experi- jelis et al. [48]. Their
past the hygroscopic range was done by Ve
mentally with a CFI-filled cavity with a density of 62.7 kg/m3
study determined moisture content of CFI in one - and two floor
where settling was not observed.
buildings with masonry walls with different thickness of insulation
For wet spray applications, the dry density of CFI has been
throughout various moisture periods measurements. A qualitative
shown to increase linearly with installed moisture content, ranging
method was used to determine the influence of moisture on vari-
from 39.6 with 40% moisture content to 71.3 kg/m3 with 100%
ations in thermal conductivity. An increase in 1% of moisture con-
moisture content [37]. If installed properly, wet spray cellulose does
tent can lead to an average increase of 1.2%e1.5% in l values for
not settle. For stabilized cellulose, an initial moist density of around
loose-fill CFI. Even when high moisture content was reached,
45 kg/m3 gets reduced to around 38 kg/m3 after drying. Settling
thermal conductivity increased from 1.6 to 2.0% for 1% of moisture
with the stabilized cellulose method in attics was found to be
content (Fig. 4). These changes in values of l are similar to those
reduced to around 5% [16].
mentioned previously by Refs. [28] and [44] in the hygroscopic
range. Generally for the hygroscopic range, the increase in thermal
3.2. Thermal properties
conductivity could be considered negligible. It is only when capil-
lary moisture begins (RH > 90%) where the insulating properties
Although the typical value for CFI's thermal conductivity is
would be ineffective. Such cases could arrive due to rain infiltration,
around 0.040 W/mK, its properties and performance can vary
leaking pipes, or improperly installed wet spray cellulose.
slightly depending on manufacturing and method of installation.
The work of Ref. [22] has shown that a difference in the source
newsprint quality can affect thermal performance. In their study,
CFI samples coming from US and Korea were measured through
heat flow meters, in accordance with ASTM C 518. By comparing CFI
from both countries, the study found that the Korean fibres that are
shorter due to having gone through more recycling processes show
a higher value for thermal conductivity, and therefore lower insu-
lating performance than CFI fibres from the US.
Since cellulose fibres are naturally hygroscopic, moisture ab-
sorption can also affect thermal conductivity values. Tye and
Spinney [46] studied loose fill CFI installed in ceiling and wall
constructions subjected to cyclic thermal and moisture gradients.
Thermal conductivity measurements were made on installed
samples using the standard ASTM C236 guarded hot box method
with a mean temperature of 15  C and a temperature difference of
approximately 10  C. It was found that thermal conductivity
increased by 15% for a moisture gain of 10%. Nicolajsen [28] found
that under the hygroscopic range (RH <90%) the change in thermal Fig. 4. Increase in thermal conductivity with moisture content of cellulose fibre
transmittance of loose fill cellulose insulation within a wall cavity insulation [48].
174 P. Lopez Hurtado et al. / Building and Environment 96 (2016) 170e177
Fig. 5. Sorption-desorption isotherm of treated cellulose insulation.
3.3. Moisture properties
The behaviour of a building material with moisture can be
determined by a series of intrinsic parameters. The sorption
isotherm of a material can determine amount of water absorbed
under different values of relative humidity. This series of values is
usually measured through continuous weighing of a cellulose
insulation sample subjected a series of changes in humidity via
saturated salt solutions.
Sorption and desorption isotherms were determined experi-
mentally by Hansen et al. [19] (Fig. 5). The isotherms are measured
at 20.0  C ± 0.5  C in a test chamber as described in EN ISO 12571, a
magnesium perchlorate solution was used as a desiccant. The dif-
ference between sorption and desorption values (hysteresis) was
negligible. Untreated CFI had a slightly lower sorption curve than
treated CFI, suggesting that the mineral additives contribute to the
adsorption of ambient humidity. A similar sorption curve was
found by Talukdar et al. [44].
Moisture diffusivity, is a property that is used in simulations to
determine the moisture concentration profile of a material. It is
defined by the moisture transport equation:
Jm ¼ r:D:grad u
With Jm ¼ moisture flux [(kg/m2s)], r the dry density of the ma-
terial [(kg/m3)], and u the moisture content [(kg/kg)]. This
parameter was determined by Marchand and Kumaran [25]. Sam-
ples of blown CFI were subjected to moisture intake and then
continuously scanned via gamma ray attenuation. These scans
provided the moisture content profiles within the material as a
function of time. Through Boltzmann transformation of these
profiles, the moisture diffusivity D was determined as a function of
moisture content within the cellulose. The value of D varied
exponentially from 5x10-8 m/s2 to 1.2x10-7 m/s2 for moisture
contents of approximately 10%e175%.
The water vapour permeability is the rate in which vapour water
is transported materials. This characteristic defines the “breath-
ability” of a material. Hansen et al. [19] determined the value of
vapour permeability of CFI from cup measurements at 23  C vary-
ing from 50% to 94% RH according to prEN ISO 12572:
177 ± 29  10 12 kg/(Pa.m.s). An increase in density (from 40 to
65 kg/m3) greatly reduced the permeability, while the removal of
mineral additives had less of an impact. The values are similar to
those found in Refs. [26,30], and [21]. Talukdar et al. [44] estab-
lished the water vapour permeability of CFI as a function of relative
humidity using ASTM Standard E96/E96M-05.
A parameter that is frequently cited by manufacturers is the
moisture buffering value (MBV) which is the ability of the materials
within the room to moderate variations in the relative humidity.
Cerolini et al. [6] calculated the MBV of CFI by exposing 69.6 g of CFI
to daily cyclic exposure of high (75%) and low (33%) relative hu-
midity levels for 8 h and 16 h. The moisture buffering value of CFI
was found to be around 3.06 [g/m2.%RH], which can be classified as
an “excellent” moisture buffer according to the scale established by
Rode et al. [36].
The highly hygroscopic nature of cellulose insulation can be
detrimental to CFI's performance, as was shown with the two
previous sections. However having a hygroscopic material in a
building envelope could theoretically be beneficial when it comes
to regulating humidity conditions inside a building, especially if a
vapour retarder is not integrated in the building envelope. Rode
[35] modelled the performance of a CFI wall under isothermal and
nonisothermal conditions in Nordic climate. In the case where no
moisture barrier or plasterboard was applied a small improvement
in interior relative humidity was found for winter months. How-
ever, external humidity conditions caused moisture accumulation
within the CFI to reach levels over 90% RH, which could potentially
promote mould growth. Hagentoft and Harderup [17] used hygro-
thermal 1D models to calculate moisture uptake of a typical wall
with a brick façade and thermal loose fill CFI insulation exposed to
Swedish climate. The study found that in when vapour retarder is
not used, moisture accumulation can reach critical levels and
possibly cause mould growth in the wooden elements of the wall.
In the work of [49], moisture transport within CFI was measured
experimentally in order to model its behaviour under massive
condensation and sub zero temperatures that create ice formation.
Their studies found that ice formation had little influence on the
water vapour permeability of the material, yet the material
continued to accumulate moisture and did not reach a steady state
within the testing period of 100 h.
Using a full scale testing chamber subjected to moisture load,
Mortensen et al. [26] found that CFI can reduce interior relative
humidity peaks by up to one half, but as with the previous studies,
this moisture reduction becomes negligible once the surface layers
of the composite wall are covered in plasterboard.
For wet spray cellulose, drying is an important factor to consider
during installation. The water from the sprayed fibres could be
(4) transmitted to wood frames cavities which could cause warping or
mould growth. A study by the Canada Mortgage and Housing
Corporation [10] found that the cellulose increased plywood
sheathing moisture content to 24% 30 days after installation, which
then reduced to 15% after 260 days. The critical moisture content in
which wood starts to develop fungi is around 30% so the moisture
Fig. 6. Calculated evolution of average moisture content in exterior facing half of wet
spray CFI with varying months of installation in Detroit Michigan [37].
 P. Lopez Hurtado et al. / Building and Environment 96 (2016) 170e177
values were acceptable. Salonvaara et al. [37] studied drying of wet
spray CFI with a hygrothermal model that takes into account the
period of installation. The study compared drying in a region with
warm dry climate vs. a region with cold humid climate. It was found
that for winter months the wet spray CFI would take many weeks to
dry and in some cases not dry at all, especially in colder region. For
example Fig. 6 shows that, during the winter months of November,
December, and January (solid blue, dashed orange, and solid black
lines in the graph, respectively) moisture content decreased by only
10% in a month. The wet spray method is therefore preferred to be
applied in warmer drier climates.
3.4. Air infiltration
Openings in a building's envelope can cause air leakage which
can have an impact on energy loads in buildings. The most cited
study by manufacturers regarding air infiltration with CFI was done
by Boonyakarn et al. [5]. The study found that the installation of CFI
(wet-spray in walls and loose-fill in attics) reduced the air change
rate from 87.5 to 29.4 ACH (Air changes per hour) with 50 Pa blower
door testing. CFI was found to improve air tightness 36% better than
fibreglass.
A study by the North American Insulation Manufacturers As-
sociation (NAHB [27] found that, via testing in wood framed walls,
wet spray cellulose had reduced the air infiltration of a structure
from 10.5 to 2.2 m3/h (converted from cubic feet per minute) while
fibreglass batts only reduced it to 6.1 m3/h. The difference in air
infiltration between types of insulation is negligible once a weather
barrier was applied.
3.5. Fire properties
The high flammability of cellulosic fibres requires them to be
treated before installation in order to achieve acceptable levels of
combustion and smouldering resistance. In a typical CFI material,
borate salts are added to prevent combustion and boric acid is
added to prevent smouldering [42]. Other additives include
aluminium sulphate, aluminium trihydrate, ammonium phosphate,
and ammonium sulphate. Day and Wiles, [13] studied the influence
of the proportions of these additives on flame spread and smoul-
dering resistance. The minimum boric acid required to prevent
smouldering as a function of borax dosage was established:
boric acid required ¼ 11:6 þ 0:185  ðborax usedÞ: (5)
Day et al. [11] found that the optimal borax/boric acid ratio of 1/
Fig. 7. Proportions (parts per hundred) of borax and boric acid in order to achieve
smoulder retardancy (///diagonal) and flame retardancy (\\\ diagonal) [12].
175
8 with a dosage 16% is necessary to prevent both flaming and
smouldering combustion (Fig. 7).
A three component formulation using borax, boric acid, and
aluminium sulphate was also studied. Varying dosage from 12% 18%
and 24% increases the possible proportions of these constituents
which allow both smouldering and combustion resistance to be
obtained. In another study (1981), they studied the effect of wetting
on additives. They establish that wetting and drying of the CFI
caused a higher concentration of both borax and boric acid to
appear on the surface of the material. This migration did not affect
smouldering resistance and would actually be favourable for flame
combustion resistance. Sprague [41] studied the consistency of
formulations and found variability in the distribution of test results.
Samples were found to attain class I or II flame resistance with a
variable distribution. As additive dosage increased, this variability
was reduced. Some of the variability was due to inconsistency in
the testing method itself.
3.6. Fungal development
It is widely known that wet lignocellulosic materials can allow
mould growth. In the case of CFI, the added additives can serve a
dual purpose of preventing mould growth as well as fire propaga-
tion. In the work of [20]; it was found that the boron included in the
cellulose was found to have a sporocidal effect on five of the most
common types of fungal spores, even when subjected to a high
concentration of fungi. For untreated fibres exposed to fungal
samples, moisture content and relative humidity was found to have
an influence on the fungal growth rate of cellulose insulation. As
the CFI samples dried, the rate of mould growth decreased.
There exist however, case studies where mould growth has been
found to be produced in houses insulated with CFI. Godish and
Godish [15] studied four wet spray CFI-insulated houses where
mould was prevalent. While the conditions in which the wet spray
CFI was applied were not detailed, (i.e. high water dosage), it was
found that two of the houses developed fungi due to rewetting of
the fibres because of water infiltration. Numerous hydrophilic
xerophilic and toxigenic species of fungus were found both within
the CFI material and in airborne samples. While this mould expo-
sure poses a risk to building occupants, properly applied wet-spray
CFI will not present these problems.
3.7. Life cycle analysis
As mentioned before, CFI has a low embodied energy compared
to traditional mineral and natural insulation materials. A compar-
ative analysis of three impact categories of the life cycle analysis
(LCA) of common insulation materials was featured in Zabalza
Bribi
an et al. [51] (see Table 3). It is worth noting that the functional
unit is 1 kg of material. Since the materials have different densities
and thermal conductivities, a more proper functional unit would be
the necessary amount of material to provide a specific value of
thermal resistance.
A more in depth LCA comparison was done by Schmidt et al.
[39], who studied the cradle to grave assessment of stone wool, flax,
and CFI, in compliance with the LCA standard ISO 14040. In this case
the functional unit was the amount of material necessary to provide
a thermal resistance of 1 m2K/W so 1.280 kg of material in the case
of loose fill CFI. The study takes into account the production of
newsprint, the manufacture of CFI, the incorporation of its addi-
tives, its installation, use and disposal for the calculation of its in-
ventory. sensitivity analysis of the end of life stage was studied. For
loose fill solutions, manufacturers state that CFI can be recycled if
no contaminants are present [4], or incinerated to provide energy
in a waste incineration plant. The study analyses the impact of
176 P. Lopez Hurtado et al. / Building and Environment 96 (2016) 170e177

Table 3
Comparative life cycle analysis of common building materials [51].

Building product density (kg/ Thermal conductivity (W/ Primary energy demand (MJ- Global warming potential Water demand (l/
m3) mK) Eq/kg) (kg CO2-Eq/kg) kg)

EPS foam slab 30 0.0375 105.486 7.336 192.729

Rock wool 60 0.04 26.393 1.511 32.384
Polyurethane rigid 30 0.032 103.782 6.788 350.982
foam
Cork slab 150 0.049 51.517 0.807 30.337
Cellulose fibre 50 0.04 10.487 1.831 20.789
Wood wool 180 0.07 20.267 0.124 2.763

partial recycling incineration or landfilling of the material. The
highest impact was caused by partial landfilling of the CFI, which
nearly tripled the global warming impact factor, due to the amount
of methane released by the material. Recycling vs incineration had
less of an impact. Interestingly, in this study CFI showed a higher
total energy consumption than stone wool(for the same functional
unit), which contradicts the studies previously shown. One reason
for this could be strategy involved in the consideration of the
impact of the manufacture of newsprint. In this study, newsprint
production represented over 90% of overall energy consumption in
the LCA of CFI. This highlights the importance of the initial hy-
potheses when analysing the life cycle of a recycled material such
as CFI.
Life cycle assessment is a useful tool in material selection for
construction projects. Takano et al. [43] studied the impact of
building material selection on the environmental characteristics of
a construction in Finland. It was found that the change from rock
wool to CFI as an insulator could reduce greenhouse gas emissions
and embodied energy of the building envelope by 15%. Similarly
[45] studied the influence of different insulation materials on pri-
mary energy and CO2 emission of a residential multi-storey build-
ing using the criteria of BBR 2012 and Passivhaus 2012 energy-
efficiency standards. It was found that the replacement of stone
wool by CFI on most parts of the building envelope resulted in
global energy reduction of 6%e7% and a decrease in CO2 emissions
from material production of 6%e8%, depending on the standard
chosen BBR 2012 or Passivhaaus). This is an interesting factor to
consider when dealing with the refurbishment of buildings, where
not only the replacement of old mineral insulation with CFI will
improve thermal properties of a refurbished building, but also
reduce its ecological impact.
4. Conclusions
As has been shown by the available literature, CFI is an inno-
vative eco-friendly insulation material that presents similar char-
acteristics in terms of thermal comfort and performance to its non-
renewable counterparts. Nevertheless the material presents some
disadvantages compared to less eco-friendly insulation materials
and has shown the need for more optimization and development.
Further research needs to focus on studying and resolving the is-
sues with the material's properties and performance.
First of all there needs to be a better understanding of the source
material. While the available research has been shown on the
performance of CFI after installation, more work needs to be done
on the manufacture and installation methods in order to further
optimize the material Studying this can be difficult due to the fact
that CFI manufacturers have many suppliers for newsprint, each of
whom may use different compositions of paper and methods of
manufacture. Nevertheless, knowing the properties of the source
newsprint in terms of paper quality, composition, fibre
morphology, and their influence on CFI insulation quality will help
reduce variability in the performance of CFI. Only the work of [22]
has compared cellulose insulation from two different sources
(Korean and US CFI), but the differences between the quality of
these two are not well detailed.
For loose fill CFI, novel methods need to be developed to reduce
settling, especially in horizontal applications. How material sepa-
ration and blowing speed during installation can affect these fac-
tors has yet to be analyzed. The stabilized approach needs to be
further developed to study the influence of water on dust exposure,
density and settling. This also applies to the wet spray method,
where the role of water dosage on the final properties of cellulose
should be investigated.
Finally changes in the formulation of CFI could be envisioned.
Other, more environmentally friendly additives with antifungal or
fire retardant properties could replace some or all of the mineral
additives used currently. The incorporation of adhesives and tack-
ifiers in the wet-spray method has been known to be used by some
manufacturers of CFI [9] but their effect on the fibres and the per-
formance of CFI has not yet been quantified.
It is through these innovations that cellulose fibre-based insu-
lations can become more prevalent and contribute to more eco-
friendly construction projects.
Acknowledgements
The work in this paper is financed by the consulting firm
Greenbuilding, as well as by grants from the French A.N.R.T (grant
gion Enterprises.
no. 2013-0149). and Paris Re
References
[1] AFNOR, NF EN 15101e1: Thermal Insulation Products for Buildings d In-situ
Formed Loose Fill Cellulose (LFCI) Products d Part 1: Specification for the
Products before Installation, 2014.
[2] M. Bomberg, C.J. Shirtliffe, Comments on standardization of density and
thermal resistance testing of cellulose fibre insulation for horizontal applica-
tions, Build. Res. Note 147 (1979) 17.
[3] M. Bomberg, K.R. Solvason, How to ensure good thermal performance of
cellulose fiber insulation. Part II. Exterior walls, J. Build. Phys. 4 (1980)
119e133, http://dx.doi.org/10.1177/109719638000400204.
[4] D.P. Boogman, EPD - Environmental Product Declaration Blown Insulation
Made of Cellulose Fibre ISOCELL GmbH, 2014.
[5] S. Boonyakarn, D. Arch, S.R. Spiezle, Fibreglass Vs Cellulose Installed Perfor-
mance, 1990.
[6] S. Cerolini, M. D'Orazio, C. Di Perna, A. Stazi, Moisture buffering capacity of
highly absorbing materials, Energy Build. 41 (2009) 164e168, http://
dx.doi.org/10.1016/j.enbuild.2008.08.006.
[7] CIMA, CIMA Technical Bulletin #2 Standard Practice for Installing Cellulose
Building Insulation, 1998.
[8] CIMA, CIMA Technical Bulletin #3 Standard Practice for Installation of Sprayed
Cellulosic Wall Cavity Insulation, 1998.
[9] CIMA, CIMA Technical Bulletin # 1: Cellulose Insulation: Codes, Regulations,
and Specifications, 1998.
[10] CMHC, Wet-sprayed Cellulose Insulation in Wood Frame Construction,
Research & Development Highlights Technical Series 90-240, Edmonton
Canada, 1990.
[11] M. Day, T. Suprunchuk, D.M. Wiles, The fire properties of cellulose insulation,
J. Build. Phys. 4 (1981) 157e170, http://dx.doi.org/10.1177/
109719638100400301.
 P. Lopez Hurtado et al. / Building and Environment 96 (2016) 170e177
[12] M. Day, T. Suprunchuk, D.M. Wiles, A combustibility study of cellulose insu-
lation, J. Build. Phys. 3 (1980) 260e271, http://dx.doi.org/10.1177/
109719638000300404.
[13] M. Day, D.M. Wiles, Combustibility of loose fiber fill cellulose insulation: the
role of borax and boric acid, J. Build. Phys. 2 (1978) 30e39, http://dx.doi.org/
10.1177/109719637800200104.
[14] European Parliament, Directive 2010/31/eu of the european parliament and of
the council of 19 May 2010 on the energy performance of buildings, 1e4,
(recast), Official J. Eur. Communities (2010).
[15] T.J. Godish, D.R. Godish, Mold infestation of wet spray-applied cellulose
insulation, J. Air Waste Manag. Assoc. 56 (2006) 90e95.
[16] R.S. Graves, D.W. Yarbrough, An Experimental Study of Stabilized Cellulosic
Insulation Installed in Four Attic Sections of Manufactured Homes, 1989.
[17] C.-E. Hagentoft, E. Harderup, Moisture conditions in a north facing wall with
cellulose loose fill insulation: constructions with and without vapor retarder
and air leakage, J. Build. Phys. 19 (1996) 228e243, http://dx.doi.org/10.1177/
109719639601900304.
[18] G. Hammond, C. Jones, Inventory of Carbon & Energy Version 2.0 (ICE V2. 0),
Department of Mechanical Engineering, University of Bath, Bath, UK, 2011.
[19] K.K. Hansen, C. Rode, E.J. Hansen, P. de, T. Padfield, F.H. Kristiansen, Experi-
mental investigation of the hygrothermal performance of insulation materials,
Perform. Exterior Envelopes Whole Build. VIII (2001).
[20] J. Herrera, Assessment of fungal growth on sodium polyborate-treated cel-
lulose insulation, J. Occup. Environ. Hyg. 2 (2005) 626e632, http://dx.doi.org/
10.1080/15459620500377667.
[21] T. Kalamees, J. Vinha, Hygrothermal calculations and laboratory tests on
timber-framed wall structures, Build. Environ. 38 (2003) 689e697, http://
dx.doi.org/10.1016/S0360-1323(02)00207-X.
[22] Y.C. Kwon, D.W. Yarbrough, A comparison of Korean cellulose insulation with
cellulose insulation manufactured in the United States of America, J. Therm.
Envelope Build. Sci. 27 (2004) 185e197, http://dx.doi.org/10.1177/
1097196304035242.
[23] M.V. Madurwar, R.V. Ralegaonkar, S.A. Mandavgane, Application of agro-
waste for sustainable construction materials: a review, Constr. Build. Mater.
38 (2013) 872e878, http://dx.doi.org/10.1016/j.conbuildmat.2012.09.011.
[24] Makron Engineering, n.d. Fibretec 1500 production line.
[25] R.G. Marchand, M.K. Kumaran, Moisture diffusivity of cellulose insulation,
J. Build. Phys. 17 (1994) 362e377, http://dx.doi.org/10.1177/
109719639401700406.
[26] L.H. Mortensen, R. Peuhkuri, C. Rode, Full scale tests of moisture buffer ca-
pacity of wall materials, in: 7 Th Nordic Symposium on Building Physics,
Reykjavík, 2005, pp. 662e669.
[27] NAHB Research Center, Air Infiltration of Wood Frame Walls, 2009.
[28] A. Nicolajsen, Thermal transmittance of a cellulose loose-fill insulation ma-
terial, Build. Environ. 40 (2005) 907e914, http://dx.doi.org/10.1016/
j.buildenv.2004.08.025.
[29] A.M. Papadopoulos, State of the art in thermal insulation materials and aims
for future developments, Energy Build. 37 (2005) 77e86, http://dx.doi.org/
10.1016/j.enbuild.2004.05.006.
[30] R. Peuhkuri, C. Rode, K.K. Hansen, Non-isothermal moisture transport through
insulation materials, Build. Environ. 43 (2008) 811e822, http://dx.doi.org/
10.1016/j.buildenv.2007.01.021.
[31] T.V. Rasmussen, Prediction of density for prevention of settling of hygroscopic
and nonhygroscopic loose-fill insulation in walls, J. Therm. Envelope Build. Sci.
28 (2005) 245e267, http://dx.doi.org/10.1177/1097196305048596.
[32] T.V. Rasmussen, Verification of density predicted for prevention of settling of
loose-fill cellulose insulation in walls, J. Therm. Envelope Build. Sci. 27 (2003)
49e69, http://dx.doi.org/10.1177/1097196303027001004.
[33] T.V. Rasmussen, Modelling settling of loose-fill insulation in walls, part II
determination of coefficients, J. Build. Phys. 25 (2002) 189e208, http://
dx.doi.org/10.1177/0075424202025003605.
177
[34] T.V. Rasmussen, Modelling settling of loose-fill insulation in walls. Part I,
J. Therm. Envelope Build. Sci. 25 (2001) 129e141, http://dx.doi.org/10.1106/
F5UH-5A2X-9VRA-FHPF.
[35] C. Rode, Organic insulation materials, the effect on indoor humidity, and the
necessity of a vapor barrier, Proceed. Therm. Perform. Exterior Envelope Build.
VII (1998) 109e121.
[36] C. Rode, R. Peuhkuri, B. Time, K. Svennberg, T. Ojanen, P. Mukhopadhyaya,
M. Kumaran, S.W. Dean, Moisture buffer value of building materials, J. ASTM
Int. 4 (2007) 100369, http://dx.doi.org/10.1520/JAI100369.
[37] M. Salonvaara, M. Pazera, A. Karagiozis, Impact of Weather on Predicting
Drying Characteristics of Spray-Applied Cellulose Insulation, 2010.
[38] P.I. Sandberg, Determination of the effects of moisture content on the thermal
transmisivity of cellulose fiber loose-fill insulation, in: Thermal Performance
of the Exterior Envelopes of Building III, ASHRAE/DOE/BTECC/CIBSE Confer-
ence 517e25, 1992.
[39] A.C. Schmidt, A.A. Jensen, A.U. Clausen, O. Kamstrup, D. Postlethwaite,
A comparative life cycle assessment of building insulation products made of
stone wool, paper wool and flax, Int. J. Life Cycle Assess. 9 (2004) 53e66.
[40] S.A. Siddiqui, Handbook on Cellulose Insulation, 1989.
[41] R.W. Sprague, Present variability in chemical add-on in the cellulose insu-
lation industry, J. Build. Phys. 2 (1979) 197e204, http://dx.doi.org/10.1177/
109719637900200403.
[42] R.W. Sprague, K.K. Shen, The use of boron products in cellulose insulation,
J. Build. Phys. 2 (1979) 161e174, http://dx.doi.org/10.1177/
109719637900200401.
[43] A. Takano, M. Hughes, S. Winter, A multidisciplinary approach to sustainable
building material selection: a case study in a Finnish context, Build. Environ.
82 (2014) 526e535, http://dx.doi.org/10.1016/j.buildenv.2014.09.026.
[44] P. Talukdar, O.F. Osanyintola, S.O. Olutimayin, C.J. Simonson, An experimental
data set for benchmarking 1-D, transient heat and moisture transfer models of
hygroscopic building materials. Part II: Experimental, numerical and analyt-
ical data, Int. J. Heat Mass Transf. 50 (2007) 4915e4926, http://dx.doi.org/
10.1016/j.ijheatmasstransfer.2007.03.025.
[45] U.Y.A. Tettey, A. Dodoo, L. Gustavsson, Effects of different insulation materials
on primary energy and CO2 emission of a multi-storey residential building,
Energy Build. 82 (2014) 369e377, http://dx.doi.org/10.1016/
j.enbuild.2014.07.009.
[46] R.P. Tye, S.C. Spinney, A study of the effects of moisture vapour on the thermal
transmittance characteristics of cellulose fibre thermal insulation, J. Build.
Phys. 2 (1979) 175e196, http://dx.doi.org/10.1177/109719637900200402.
[47] UNEP, Buildings and Climate Change: Status, Challenges, and Opportunities,
UNEP DTIE, Sustainable Consumption and Production Branch, Nairobi, Kenya ,
2007 (United Nations Development Programme ;; Paris, France).
[48] S. Vejelis, I. Gnipas, V. Kersulis, Performance of loose-fill cellulose insulation,

Mater. Sci., (MEDZIAGOTYRA) 12 (2006).
[49] T. Vrana, K. Gudmundsson, Comparison of fibrous insulations e cellulose and
stone wool in terms of moisture properties resulting from condensation and
ice formation, Constr. Build. Mater. 24 (2010) 1151e1157, http://dx.doi.org/
10.1016/j.conbuildmat.2009.12.026.
[50] B. Wu, C.M. Taylor, D.R.U. Knappe, M.A. Nanny, M.A. Barlaz, Factors controlling
alkylbenzene sorption to municipal solid waste, Environ. Sci. Technol. 35
(2001) 4569e4576, http://dx.doi.org/10.1021/es010893a.
[51] I. Zabalza Bribia n, A. Valero Capilla, A. Aranda Uson, Life cycle assessment of
building materials: Comparative analysis of energy and environmental im-
pacts and evaluation of the eco-efficiency improvement potential, Build. En-
viron. 46 (2011) 1133e1140, http://dx.doi.org/10.1016/
j.buildenv.2010.12.002.
[52] R. Zaborniak, Prediction of long-term density of cellulose fibre insulation in
horizontal spaces of new residential houses, J. Build. Phys. 13 (1989) 21e32,
http://dx.doi.org/10.1177/109719638901300104.